This empirical paper applies scanner panel data to estimate brand choice price-elasticities for two groups of households: loyal and non-loyal consumers to a brand of a consumer good. The main objective is to verify differences in price sensitiveness among those consumers when they choose brands. The method of Maximum Likelihood and the approach of utility maximization are used to estimate the mixed logit model with two types of parameter vectors: household characteristics and brand attributes. The results show that differences in price elasticities among consumers vary according to each brand, implying that marketing managers can take more effective decisions when considering household characteristics and brand attributes. Such analysis can be helpful for the Brazilian retail sector, as the availability of scanner data has been increasing since last decade.
price-elasticity; logit model; scanner data; consumer choice
